does anyone know if can 3DMark be affected by faulty system clock?

[shiyan]

Just wondering if anyone knows if 3DMark's scores can be affected by an inacurate system clock...

[f00t]

I would be guessing not too much. if the cpu were ff by about 20mhz than you might loose a point or two

[PreservedSwine]

you mean the system clock that keeps time, not the cpu clock, right?

If that's what you mean, then no.

Of course the cpu clock will affect the score, but not the little time clock....

[shiyan]

yeah, I meant the clock that keeps time.

what I was thinking of was whether a clock that was too fast or too slow might influence the score, but I guess not then?

[UnseenMenace]

Simply put yes it can be... This was my old Duron 800 a while back (notice the CPU speed, pick a benchmark score you would like... :p )

http://www.overclockers.com/articles444/madonion.jpg

This benchmark is also effected by running under emulation in non Windows OS's and in certain ways using virtual machines.

Why people believe a benchmark that is produced with the assistence of the graphic card companys is beyond me anyway... its not exactly independant is it

[PreservedSwine]

He's not talking about his cpu clock.

He's talking about the time clock powered by his BIOS battery. It has NO effect on benchmarking:)

[Kermit_the_Frog]

If his clock wont keep time I would think change the battery, and maybe look for a bios update.


:cool:

[CamH]

Let's see those scores, UnseenMenace. :D

[TR_B]

3dm score -> Frame Per Second

How accurate would every pc "know" its time?
For all I know one second of time on my pc, is 0.98 second on someone elses.

[UnseenMenace]

Originally posted by CamH
Let's see those scores, UnseenMenace. :D

http://www.overclockers.com/articles444/Sandracpu.jpg

Not bad for a Duron 800 on a Abit KT7A-Raid with Crucial PC133 back in the day :p

[UnseenMenace]

The SI Sandra score in my previous post was created two years ago using a game cheat that effects the computers perception of time.. in the game it makes the other players move slowly so that you can shoot them easier... This one only effects certain benchmarks, however there are other game hacks, applications and techniques available to someone who wishes to fake a different benchmark
I could easily post benchmarks which have been altered by Game Hacks, Applications, or emulation, or by running them in Virtual machines or liteOS... and as before someone would suggest that they had been photoshopped or came from a more powerfull PC than claimed..

Personally I believe there are Facts, BS and Benchmarks... two of these I take with a pinch of salt... The information is out there do the research and believe what you wish

I did a article on the subject of benchmarks being effected by the clock for the front of overclockers.com two years ago when my knowlege and understanding of such things was of a lower standard.. Upon reflection the article has some major gaps and could be expanded so much futher.. However it served its purpose in making people question benchmarks in general.
I recieved a lot of hurtfull flame mail and problems from the article when it was written, yet to this very day I remain exceptionally interested in benchmarks... however the responce I recieved from the article makes me reluctant to disclose my more recent findings or discuss how my thoughts regarding benchmarks have developed or changed over this time.

http://www.overclockers.com/articles444/

[StormPC]

Originally posted by shiyan
Just wondering if anyone knows if 3DMark's scores can be affected by an inacurate system clock...

NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO!:D

[shutoutkeeper3]

I think it could affect it... I mean think about it if it thought a second was like 2 or 3 seconds then it would think you are getting twice the FPS you are... Kinda related, I once heard a story about a teacher overclocking his PDA (i don't know how) and used his PDA clock to time tests. Since he had overclocked the PDA his clock went to fast, so the students got less time on the tests. Once he found out what had happened he gave the students back the lost time...

[StormPC]

Originally posted by shutoutkeeper3
I think it could affect it... I mean think about it if it thought a second was like 2 or 3 seconds then it would think you are getting twice the FPS you are... Kinda related, I once heard a story about a teacher overclocking his PDA (i don't know how) and used his PDA clock to time tests. Since he had overclocked the PDA his clock went to fast, so the students got less time on the tests. Once he found out what had happened he gave the students back the lost time...

Uh, nope! Benching has nothing to do with the RTC. System clocks are not based on the RTC. Clock generators determine CPU and buss clocks, not the RTC.:o

In the case you spoke about the PDA does not have a RTC, so the time was generated by the system clock generators which is not an accurate way to do it, but it's cheaper than including an RTC in the PDA.

[shutoutkeeper3]

Originally posted by StormPC


In the case you spoke about the PDA does not have a RTC, so the time was generated by the system clock generators which is not an accurate way to do it, but it's cheaper than including an RTC in the PDA.

Well it's accurate in normal conditions... PDA's aren't meant to be overclocked... That's what made the story funny... :p

[StormPC]

Originally posted by shutoutkeeper3


Well it's accurate in normal conditions... PDA's aren't meant to be overclocked... That's what made the story funny... :p

No computer is "meant" to be overclocked, but some have overclocking features. PDA's are no different than desktops in that respect. I've overclocked a couple of PDA's because the manufacturer allowed it.:cool: